What ccusage gets right
ccusage is a good free CLI: run it with npx and it reads the usage records coding agents keep on your machine, then prints daily, weekly, monthly, and per-session tables with cost estimates. It also offers active monitoring for Claude Code 5-hour blocks and a Claude status-line integration documented as beta at the time of writing. It covers many agent CLIs and is open source (MIT-licensed at the time of writing).
A native control surface is a different job
AgentPeek keeps live gauges in the Mac notch and menu bar: 5-hour and 7-day windows for Claude Code, the 7-day window for Codex with reset countdowns, Cursor's monthly window, and token totals for local-record agents. Unlike a terminal report or Claude status line, these native surfaces sit beside session operations.
- Budget alerts fire once at your threshold, before a limit interrupts a refactor.
- AgentPeek shows only the limit windows an agent itself reports.
- A by-day chart and input, output, and cache split cover the history questions in the same panel.
Usage is one page of it
AgentPeek treats usage as one surface among many. The same app shows every live session with its state, plus the transcript, diffs, and todos its source exposes, surfaces permission prompts you can answer with a keystroke, sends the next prompt into a resumable session, sorts everything on the Agent Board, and lists your local dev servers. A usage CLI is one tab of that job.
Run both if you like
This is not either-or. Their local usage readers do not mutate the records they inspect, so the tools can run side by side: keep ccusage for scripted exports, terminal reports, and its Claude block view, then use AgentPeek for native session monitoring and controls. The build vs buy page compares AgentPeek's fixed license price with the scope you would otherwise maintain.
Feature snapshot as of August 2026; check each project's repository for current features and license.
